INDEX

Explanations

`&` followed by html entities or commands

np_acts-logits-general · gemini-2.5-flash-lite

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&” or “>”).

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 específica

-1.80

 japonesa

-1.70

 avanzada

-1.62

 trám

-1.59

 Тогда

-1.56

 Затем

-1.54

 enchufe

-1.54

 underpin

-1.54



-1.53

恸

-1.52

POSITIVE LOGITS

all

1.84

ｕ

1.81

裏面

1.70

ｋ

1.68

)++;

1.66

可以说是

1.61

ｎ

1.61

 Toscana

1.60

ですから

1.59

無い

1.55

Activations Density 0.001%

`&` followed by html entities or commands

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&” or “>”).

No Comments

No Known Activations

`&` followed by html entities or commands

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&” or “>”).

No Comments

No Known Activations

`&` followed by html entities or commands

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&amp;” or “&gt;”).

No Comments

No Known Activations

`&` followed by html entities or commands

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&amp;” or “&gt;”).

No Comments

No Known Activations

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&” or “>”).

The neuron fires on the semicolon that terminates HTML‐style character entities (the “;” in sequences like “&” or “>”).